350 Trowbridge and McRae—Elasticity of Ice. 
gave the elongation for a given weight. The increased length 
being so small and the least wind causing the fiber and bar to 
swing, the errors of observation were so large that this 
method was abandoned. 
It may be interesting to note that a bar left suspended an 
entire day (temperature —18° C.) changed from a cylinder to 4 
frustrum of a cone. 
vibrations of the e bars were obtained in the same 
manner as in the previous method. ewing needle eee 
comparing the number in equal distances along the two ab 
it was clamped or when it was struck to set it in vibration, but 
by continued trials we succeeded in obtaining two well-detin 
curves. The height of the vibrations was about one-third wal . 
millimeter. It is not certain that the ice always attained 1 
proper pitch. ; 
he results are expressed in the following formula: 
ee n2l4g 
~ (28) 2r2 ‘ae 
in which /= length of the bar, r= the radius, s= Ae ns 
gravity and n= the number of transverse vibrations per § 
